This research attempts to garner a better understanding of the ²nature and dynamics of the microbial habitats″ within an artificial wetland treating domestic sewage. The wetland studied is a top loading vertical flow submerged bed system (Figure 1) located in Highland, NY, USA. The treatment wetland consists of four, 237-m² beds, which may be operated in parallel (two trains of two beds in series) or as four beds in series. Two beds are planted with Reed Canary Grass (Phalaris arundinacae) and two with common reed (Phragmites sp.). Current operation is four beds in series with a mean flow rate of 7.5 m³ d^-1. The treatment system is subject to a NYS NPDES permit. Permit levels are: BOD₅, 5 mg L^-1; TSS, 10 mg L^-1; NH₄⁺, 2.2 mg L^-1; and PO₄^3-, <1 mg L^-1; respectively. Redox potentials are measured with platinum tipped probes installed at 15, 30, and either 45 or 60-cm-depth within each bed. Temperature, pH, dissolved oxygen and precipitation data are also measured on site. System influent and effluent samples from each of the four treatment units are collected, iced and returned to the laboratory for analysis. Sand samples are collected at five locations within each treatment unit, for microbial measurements, which include: measurement of heterotrophic potential (1-¹⁴C-D-Glucose addition); denitrifier enzyme activity (acetylene blockage); nitrification potential (chlorate blockage); and microbial biomass (simultaneous chloroform exposure and extraction). Plants were harvested (fall 2005) from three locations within each treatment unit, dried and analyzed for total biomass and biomass carbon and nitrogen. Mean influent wastewater values over a 2-year period have been 293.7 mg L^-1, 72.1 mg L^-1, 116.2 mg L^-1 and 16.5 mg L^-1 for BOD₅, TSS, NH₄⁺ and PO₄^3-, respectively. Mean effluent values for BOD₅ and TSS have been <4 mg L^-1 and <1 mg L^-1, respectively; these levels are achieved in the first bed. Heterotrophic potential, however, is significantly lower in the first bed 4.8 μg g^-1 h^-1, with a steady increase through the next three beds, 11.5, 13.3, and 16.2 μg g^-1 h^-1, respectively. Eh values at the 45-cm-depth are lower in the fourth bed (second Phragmites bed) than in the second bed (second Phalaris bed), with mean Eh values of 65.8 ± 58.4 mV and 295.1 ± 67.4 mV, respectively. It appears that sufficient oxygen is present, in a number of microcosms, to reduce the ammonium from an influent mean concentration of 116.2 mg L^-1 to non-detectable concentrations leaving the third and fourth beds. Periodic increases in effluent nitrate concentration (10 mg L^-1 to 20+ mg L^-1), however, are followed in time (approximately 7 days) by similar increases in effluent nitrite concentration (20 μg L^-1 to 200 μg L^-1) in the first and third beds of the system. Additional experiments are currently being undertaken to determine the possible causes for the apparent disturbances of the Nitrobacter sp. populations. Figure 1. Schematic cross section of a top loading vertical flow submerged bed treatment wetland.